The Ding Laboratory

Current Research Areas:

    ● Stem Cell Biology and Regeneration.
        • Directed Differentiation of Pluripotent ES Cells.

        • Human Embryonic Stem Cells.
        • Directed Differentiation of Multipotent Mesenchymal Stem Cells.
        • Directed Differentiation of Neural Stem Cells.
        • Cellular Dedifferentiation.
        • Developmental Pathway Screens.
        • Exploring Regeneration in Mice.

    ● Chemical and Functional Genomic Technologies.
        • The chemical libraries.
        • The Genome-wide cDNA and siRNA Libraries.
 

Stem Cell Biology and Regeneration.

        Mankind has been fascinated with regeneration of body parts for thousands of years. It was dated back to the early 18th century that Abraham Trembley carried out the first scientific study of hydra regeneration. In 1901, Thomas H. Morgan stated in his book that "The regenerative process is one of the fundamental attributes of living things...". Since then, stimulating true (epimorphic) regeneration of damaged tissues or organs in mammals has been a holy grail. Regeneration usually involves either (1) epimorphosis where replacement cells arise from undifferentiated cells (either stem cells or dedifferentiated cells) that form a blastema from which the new structures derive; and/or (2) morphallaxis where new cells are derived from existing tissues by cell differentiation and/or migration. Among organisms that have regenerative capabilities, the urodele amphibians are unique in their use of a cellular dedifferentiation mechanism at the damaged site to form a blastema which contains dedifferentiated progenitor cells. These cells can proliferate and re-differentiate under developmental control to regenerate a wide variety of tissues, such as limbs, tails and lens.  In mammals, epimorphic regeneration is largely limited by an irreversible differentiation process. My laboratory is interested in studying the molecular mechanism of regeneration at both cellular and whole organism levels and developing new strategies to induce regeneration in vivo.
       

Small Molecules and Stem Cells.

Stem cells are unspecialized precursor cells with the ability to self-renew and differentiate into specialized cells in response to appropriate signals. Stem cell fate is controlled by both intrinsic regulators and the extra-cellular environment (niche), and is typically controlled ex vivo by cell culture manipulation with “cocktails” of growth factors, signaling molecules, and/or by genetic manipulation. Cell-based phenotypic assays and, more recently, pathway screens of synthetic small molecules and natural products have historically provided useful chemical tools to modulate and/or study complex cellular processes. Cell permeable small molecules such as dexamethasone (a glucocorticoid receptor agonist), ascorbic acid, 5-azacytidine (5-aza-C, a DNA demethylating agent), and all-trans retinoic acid (RA) have proven extremely useful for inducing the differentiation of various stem cells (e.g., embryonic, neural and mesenchymal stem cells).

Directed Differentiation of Pluripotent ES Cells. Pluripotent embryonic stem (ES) cells represent a potential unlimited source for all types of specialized cells and are ideally suited for studies of differentiation and early embryonic development. The most general method to differentiate ES cells is to grow them as aggregates in suspension without LIF to form embryoid bodies, which spontaneously differentiate into various cell lineages. The differentiation programs of EBs can be shifted toward cardiomyogenic, neuronal, or primordial germ cells by treatment with small molecules, such as dimethyl sulfoxide (DMSO) or RA, at specific stages of differentiation. However, this approach is not very efficient, and normally requires selection to enrich specific cell lineages.

By screening our combinatorial libraries of heterocyclic compounds (see in the chemical library section), we have discovered small molecules that can specifically control stem cell fate in various systems. For example, using an engineered P19 reporter cell line (stably transfected with the pTα1(tubulin)-luciferase construct) for the primary high throughput screen, we discovered a 4,6-disubstituted pyrrolopyrimidine, TWS119, that can induce neuronal differentiation of pluripotent murine EC and ES cells. We further identified one target of TWS119 is glycogen synthase kinase-3b (GSK-3b) by both affinity-based (using TWS119-linked agarose affinity matrix) and biochemical methods (using surface plasmon resonance and kinase inhibition assays). Similarly using reporter and imaging assays to screen the large chemical libraries, we discovered a series of 2,4-diaminopyrimidine analogs, cardiogenols, that can specifically induce cardiomyogenesis of mouse EC and ES cells into beating cardiac muscle cells. We are currently continuing our efforts on identifying potent and selective small molecules which can direct murine and human ES cells toward neuronal cells (subtype specific), cardiomyocytes, insulin secreting β cells, and germ cells, as well as characterizing their mechanism of action by various approaches.

      Human Embryonic Stem Cells. The basic biology and therapeutic values of ES cells are ultimately required to be investigated in the human system. However, there are significant differences between mouse and human ESCs in terms of self-renewal and differentiations. Thus, basic questions about mouse and human ESCs should be addressed in parallel. We have obtained several human ES cell lines from both WiCell Research Institute at University of Wisconsin and Dr. Melton's lab at Harvard. Currently, we are studying these hESCs by using small molecule libraries, arrayed cDNA and RNAi libraries, as well as carrying out functional genomic and proteomic studies toward finding small molecules and genes that can control self-renewal and differentiations of human ESCs to functional cells in various tissue types.

       Directed Differentiation of Multipotent Mesenchymal Stem Cells. Mesenchymal stem cells (MSCs) are multipotent progenitor cells located in the bone marrow that can differentiate into a variety of non-hematopoietic tissues such as osteoblasts, adipocytes and chondrocytes. Previous studies have shown that adult MSCs, when injected into animals, are capable of homing to a site of injury and restoring tissue function.

        To identify small molecules that selectively differentiate MSCs into defined lineage-committed progenitor cells, we screened our combinatorial heterocyclic compound library in the mouse mesenchymal progenitor cells (C3H10T1/2 cell line) for molecules that induce osteogenesis. A high throughput fluorescence-based enzymatic assay was used to detect the bone specific marker, alkaline phosphatase (ALP). A 2,6,9-trisubstituted purine compound, purmorphamine, was identified as a potent osteoblast differentiation inducing agent. Purmorphamine can activate Cbfa1/Runx2 (a master regulator of bone development), as well as upregulate other bone specific markers, such as osteopontin and collagen-I, and cells treated with purmorphamine have an osteoblast morphology. This molecule also displays phenotypically distinct osteogenesis inducing activity relative to BMP-4. Purmorphamine also showed a synergistic effect with BMP-4 in inducing osteogenesis of C3H10T1/2 cells, as well as transdifferentiating preadipocytes (3T3L1) and myoblasts (C2C12) to osteoblasts. Genome-wide mRNA expression profiling revealed that purmorphamine acts through the Hh pathway. In addition to directed osteogenesis of MSCs, we are also taking steps toward finding small molecules which can selectively induce myogenesis of MSCs (since the discovery of myogenic master genes more than 20 years ago via 5-azacytidine, which does not directly activate a specific differentiation program, but rather converts the cells into a competent spontaneous differentiation state, selectively inducing myogenesis by small molecules has been extremely challenging).

        Directed Differentiation of Neural Stem Cells. The recent discoveries of neural stem cells (NSC) in the adult central nervous system (CNS) and their regenerative roles in brain damage may make possible new approaches to the treatment of neurodegenerative disease and CNS injury. These could involve cell replacement therapy and/or drug treatment to stimulate the body’s own regenerative mechanisms by promoting survival, migration, proliferation, and/or differentiation of endogenous CNS stem cells. However, such approaches will require the identification of renewable cell sources of graftable functional neurons, and an improved understanding of and ability to manipulate neuronal development.

        High-content imaging in conjunction with immunofluorescent labeling of neuronal (bIII tubulin, TuJ1) and astroglial markers (glial fibrillary acidic protein, GFAP) has been used to screen our chemical libraries for molecules that can direct the differentiation of primary NSCs isolated from adult rat hippocampus specifically into neurons or astroglia. Specifically, treatment of adult rat primary NSCs in monolayer with aminothiazole compound (which was discovered from the screens) for four days induced up to 80% cells to differentiate into TuJ1 positive neurons with the characteristic neuronal morphology. Expression of the neuronal bHLH transcription factors NeuroD1 was also shown to be upregulated, and expression of Sox2 (a neural progenitor marker) decreased after compound treatment. Importantly, the compound was also shown to suppress astroglial differentiation induced by BMP2 and LIF combination treatment, while RA failed to do so. We are currently examining the in vivo effects of this molecule. Other ongoing projects include identifying small molecules which can promote proliferation of NSCs, as well as selectively induce differentiation of subtype-specific neurons (such as dopaminergic neurons).

Cellular Dedifferentiation. A long-standing notion in developmental biology has been that organ/tissue-specific stem cells are restricted to differentiating into cell types of the tissue in which they reside. In other words, they have irreversibly lost the capacity to generate other cell types in the body. Indeed, recent studies have demonstrated that the in vivo “plasticity” of adult stem cells is largely attributed to cell fusion events. On the other hand, a number of reports have demonstrated that tissue-specific stem cells may overcome their intrinsic lineage-restriction upon exposure to a specific set of in vitro culture conditions, although this reprogramming does not reflect potentials that are normally exercised in vivo. An extreme example is the reprogramming of a somatic cell to a totipotent state by nuclear transfer cloning, where the nucleus of a somatic cell is either transferred into an enucleated oocyte, or the extracts of the oocyte are fused with a somatic cell. Although in mammals neither transdifferentiation nor dedifferentiation has yet been identified as a naturally occurring process (except certain disease states), the discovery of stem cell plasticity raises the possibility of reprogramming restricted cell fate. The ability to dedifferentiate or reverse a lineage-committed cells to multipotent progenitor cells might overcome many of the obstacles associated with using ESCs and adult stem cells in clinical applications (inefficient differentiation, rejection of allogenic cells, efficient isolation and expansion, etc.). With an efficient dedifferentiation process, it is conceivable that healthy, abundant and easily accessible adult cells could be used to generate different types of functional cells for the repair of damaged tissues and organs.

To identify small molecules that induce true dedifferentiation, we designed a screen based on the notion that lineage-reversed myoblasts should regain multipotency. Specifically, dedifferentiated myoblasts should acquire the ability to differentiate into multiple non-permitted cell lineages when exposed to conditions that typically induce differentiation of multipotent mesenchymal progenitor cells into adipocytes, osteoblasts or chondrocytes. Dedifferentiation was initially coupled to osteogenic differentiation for the primary screen since there are an established osteogenic inducing condition and an established high throughput assay for detecting the bone specific marker, ALP. A two stage screening protocol was used in which C2C12 cells were treated with compound in the combinatorial heterocycle library for four days to induce dedifferentiation, compound was then removed and cells were assayed for their ability to undergo osteogenesis upon addition of known osteogenic inducing agents (e.g. 50 μg/ml ascorbic acid 2-phosphate, 0.1 μM dexamethasone and 10 mM β-glycerophosphate), which have such effect only on mesenchymal progenitor cells. Among a series of 2, 6-disubstituted purine analogs identified in the primary screen, 2-(4-morpholinoanilino)-6-cyclohexylamino-purine (named reversine) was found to have dedifferentiation activity. Consistent with this observation, reversine inhibits myotube formation and treated myoblasts continue to grow to form a confluent culture of mononucleated cells, which can redifferentiate into osteoblasts and adipocytes upon exposure to appropriate differentiation conditions. We are currently attempting to further improve the compound's activity by generating new analogs, and characterize the cellular effects of the molecule as well as identify the molecular basis for its activity using both affinity-based and functional genomic tools. In collaboration with Dr. Hans Schöler at MPI, we are also working toward finding molecules which can dedifferentiate human and mouse fibroblasts back to pluripotent embrynic-stem-cell like cells.

       Developmental Pathway Screens. In parallel to directly studying stem cell proliferation and differentiation, we are also carrying out pathway-based screens (small molecule screens using chemical libraries and genetic screens using arrayed cDNA and siRNA libraries) to specifically interrogate Wnt, Hh, Notch and BMP signalings. Small molecules and genes discovered through these screens are currently/will be followed up biochemically and functionally in various stem cell systems.

        Exploring Regeneration in Mice. Ear hole punch is a widely used technique to mark mice in animal studies, as such ear-punched hole will never close. Serendipitously a strain of MRL mouse was discovered that can heal 2-mm ear hole without scarring after four weeks. This mouse also exhibits rapid optical nerve and heart regeneration. However, efforts toward mapping the gene loci that confer these unique regeneration properties largely failed because of its complicated genetic background. Therefore, we have carried out large scale systematic ENU mutagenesis screen for healer mice using the ear-hole-punch model. We have already identified three families of mutant mice that showed inheritable regenerative wound healing/hole closure. Currently, continued ENU screen (ear-hole closure and lens regeneration), additional characterizations and gene mapping are underway with the goal of ultimately identifying regeneration factors.

Chemical and Functional Genomic Technologies:

        The chemical libraries. One approach to the generation of functional small molecules that control stem cell fate involves the use of cell-based phenotypic or pathway-specific screens of synthetic chemical or natural product libraries. With recent advances in automation and detection technologies, millions of discrete compounds can be screened rapidly and cost-effectively. However, although combinatorial technologies allow the synthesis of a large number of molecules with immense structural diversity, it is impossible to saturate chemical space. Consequently, the careful design of chemical libraries becomes a critical aspect of combinatorial synthesis. Ideally, the properties of a chemical library are optimized to interact with the specific biomolecules or collection of molecules of interest.  Previously, we developed the concept of using molecular scaffolds themselves as a diversity element for combinatorial synthesis. In this approach, a variety of naturally occurring and synthetic heterocycles that are known to interact with proteins involved in cell signaling (e.g., kinases, cell surface receptors, etc.) were used as the core molecular scaffold. These included substituted purines, pyrimidines, indoles, quinazolines, pyrazines, pyrrolopyrimidines, pyrazolopyrimidines, phthalazines, pyridazines, and quinoxalines. General synthetic schemes (Scheme on the right) were then developed that could be used in parallel reactions to introduce a variety of substitutents into each of these scaffolds to create diverse chemical libraries. For example, a second diversity element could be introduced into these heterocyclic scaffolds using solution-phase alkylation or acylation reactions. This was followed by capture of the modified heterocycles onto solid support using different immobilized amines to introduce a third diversity element. The resin-bound heterocycles could then be further modified (introducing a fourth diversity element) through a variety of chemistries including acylation, amination and palladium-mediated cross coupling reactions with amines, anilines, phenols, and boronic acids.

Using these chemistries in conjunction with the “directed-sorting” method, we have generated a privileged heterocycle library consisting of over 35 distinct structural classes and 100,000 discrete small molecules, which has proven to be a rich source of biologically active small molecules.

        The Genome-wide cDNA and siRNA Libraries. Genome-wide methods for directly and functionally annotating gene functions in a given biological system/assay are highly desirable. Conventional functional screens have been performed largely on libraries of pooled cDNA clones. Because of the complexity of such libraries and the need to over-sample them to find rare clones, it is necessary to screen large numbers of clones to identify those with a desired activity, limiting assays with exceedingly simple readouts, such as growth selection, and subsequent deconvolution of clone identities. Armed with high throughput screening platform, the use of individually arrayed cDNA library, whose each member is spatially separated in different wells of multi-well plates, enables performing more complex screens such as monitoring dramatic morphological changes during differentiation of stem cells as well as assays detecting fast dynamic changes without the need for clone rescue and deconvolution (since clone identity is ascertained by the well location).

siRNA Expression Libraries. In parallel to the concept of over-expressing arrayed cDNAs for gain-of-function study, this approach can also be applied to arrayed small interfering RNA (siRNA) library for loss-of-function study. Post-transcriptional gene silencing (PTGS) or RNA interference (RNAi) is the phenomenon in which gene expression is suppressed via mRNA degradation with the introduction of a homologous double-stranded RNA (dsRNA). Recently it has been demonstrated that double-stranded siRNA with 21 nucleotides (nt) plus 2 nt 3' overhangs can induce efficient gene silencing in mammalian systems. Such siRNA molecules can be transcribed in cells from transfected siRNA expression vectors or cassettes containing RNA pol III promoters.

 To facilitate the construction of large genome-wide libraries of small interfering RNAs (siRNA), we have developed a dual promoter system (pDual) in which a synthetic DNA encoding a gene-specific siRNA sequence is inserted between two different opposing Pol III promoters, the mouse U6 and human H1 promoters. Upon transfection into mammalian cells, the sense and antisense strands of the duplex are transcribed by these two opposing promoters from the same template, resulting in a siRNA duplex with a uridine overhang on each 3' terminus, similar to the siRNA generated by Dicer. These siRNAs can be incorporated into the RISC without any further modification. A single step PCR protocol has been developed using this dual promoter system which allows the production of siRNA expression cassettes in a high throughput manner. We have shown that siRNAs transcribed by either the dual promoter vector or siRNA expression cassettes can induce strong and specific suppression of both endogenous genes and ectopically expressed genes in mammalian cells. This single-step PCR approach using one gene-specific oligonucleotide is efficient and cost effective, and provides a high throughput and practical method for generating large libraries of gene specific siRNAs for genome-wide loss-of-function cellular screens. Thus, we constructed large siRNA expression cassette libraries targeting over 16,000 human and mouse genes with more than two targeting sequences designed per gene based on proprietary algorithm). These genes represent a large fraction of the proteins known to be involved in biosynthesis, metabolism, signal transduction, gene regulation and cell cycle control; the distribution of these genes is shown below.

         Arrayed cDNA Libraries. In collaboration with GNF, arrayed cDNA libraries containing over 30,000 full-length human and mouse cDNA clones from Mammalian Gene Collection (MGC, mgc.nci.nih.gov) and OriGene Technologies (www.origene.com) are available for various cellular screens.